Chiral coordination complexes are frequently used in asymmetric synthesis and chiral discrimination technologies (Eliel et al., 1994). Recently, C3-symmetric chiral ligands have shown great potential for enantioselective reactions, yet few such compounds are available (Burk et al., 1990).
A recent and exciting prospect in the area of information technology lies in the development of molecular switches that operate with efficiency, reversibility, and resistance to fatigue (Lehn, 1995). The development of electrochemical switches has recently attached much attention due to applications such as data storage (Huck et al., 1996).
Redox switches require:
(A) components whose structures and physical properties can be turned on or off electrochemically (Fabbrizzi et al., 1995); and
(b) sufficiently different optical spectra that allow the individual states to be addressed (Huck et al.).
(Feringa et al., 1991) have used isomerism in thioxanthenes to obtain chiroptical switches. Lights of different wavelengths was used to switch between M and P isomers. The difference in chirality/helicity leads to a different response in the circular dichroism (CD), where Δε=45(M) vs Δε=49(P)). The enantiomers 1 and 2 in FIG. 1 were found to be stable at room temperature. Thermal racemization of isomer 1a showed first-order reaction kinetics with a barrier to racemization of 26.4 kcal/mol. Cis-trans isomerization was not observed under ambient conditions. On the other hand, irradiation of pure 1a, presumably at ambient temperature, yielded 64% 1a and 36% 2a, with no observable racemization, while irradiation at 250 nm gave 68% 1a and 32% 2a. Furthermore, alternate irradiation of 1a at 250 and 300 nm caused a modulation of the circular dichroism (CD) signal-for 1a at 232 and 262 nm for a switching time of three seconds. Similar behavior was observed for switching times of 0.5 to 60 seconds. Compound 1 could be cycled between cis and trans forms a minimum of 10 times without racemization of changes in the UV and CD spectra. Feringa added his racemate mix to a nematic liquid crystal and then irradiated them with circularly polarized light. The resulting excess of one enantiomer was enough to switch the liquid crystal into its chiral cholesteric phase.
Shining ordinary light on the mix will convert the liquid crystal and the racemate back again (Feringa et al., op. cit.). Important problems that remain to be solved are improved thermal stability, increased fatigue resistance, and structural modification to achieve switching with visible light.
A different approach towards an optical switch was published by Schuster et al. Schuster recognized that in an opto-optical switch, where the position of the switch is transposed with light and sensed with light by change in its absorption spectrum, a fundamental problem is the destructive readout. That is, reading the position of the switch ultimately erases it. To overcome this potential problem, Schuster developed a system consisting of a photochromic fulgide dissolved in a photochemically inert cholesteric liquid crystal that is bistable and switchable by repetitive application of ultraviolet and visible light. In the course of the experiment, he showed that the pitch of a cholesteric liquid crystal can be controlled photochemically by the photochromic fulgide dopant. The pitch changes were measurable at reasonable fulgide concentrations, and both states of the liquid crystal/fulgide mixture are thermally stable under the conditions tested. The change of the pitch was bi-directional and reversible, and is controlled with light of suitable wavelength. According to Schuster, the pitch of the liquid crystal can be read optically without affecting the record information (Janicki et al., 1995). This is shown in FIG. 2.
Daub et al. reported a dihydroazulene/vinylheptafulvene photochroism system in which the information is stored and read with light. Besides the disadvantage of both storing and reading with light, the system possesses several chemical modification sites that might permit the tuning and optimization of the switching behavior (Spreitzer et al., 1996). This system is shown in FIG. 3.
Coordination complexes featuring iron in its two distinct oxidation states embedded in a triple stranded ligand were inter-converted by chemical means. This system took advantage of a “hard” binding cavity and a “soft” binding cavity present within the system. The iron literally translocated within the strand depending upon its oxidation state. The process was monitored by the UV (d-d transitions of the Fe(II) as well as the Fe(III) species). The system did not display reversible behavior. In fact, oxidation of the Fe(II) species had to be facilitated at 50° C. in order to obtain the Fe(III) complex (Zelikovich et al., 1995).
Another version of this redox switch inter-converts between two distinct states by ligand exchange. At the heart of this switch is a molecule that possesses two sets of binding groups: one set of hard and one set of soft ligating groups. The two sets are anchored on a calix[4]arene ring in an alternating fashion, such that they can form either a hard or a soft ion biding cavity. One cavity is formed at the exclusion of the other, according to the authors. When loaded with Fe(III), the hard binding groups, hydroxamates, converge to embrace the hard metal ion, while soft groups diverge. Upon reduction, the ligand rearranges to engulf Fe(II) with its soft bipyridyl groups, while the hard groups diverge. Subsequent oxidation reversed the process. This switch action was again followed by UV (Canavet et al., 1996). Stoddart et al. developed a synthetic methodology based on the idea of assembling carefully designed small molecular components in a template-directed manner. The molecular subunits are not held together by classical covalent bonds, but rather by twinning and interlocking, the mechanical interactions responsible for the presence of catenanes, rotaxanes, and knots. They described a synthesis of catenanes and a rotaxane which were able to function as “molecular trains” and a “molecular shuttle.” As seen in FIG. 6, the [2]-rotaxane 1, which can operate as a molecular shuttle, consists of a molecular assembly in which a tetracationic bis-pyridinium cyclophane moves back and forth like a shuttle (1<—>2) between two “stations” which are situated symmetrically in a polyether terminated at the ends by large groups that acct as “stoppers.” The positively charged cyclophane ring will be attracted equally by the two identical electron-rich hydroquinol groups and therefore jump back and forth between the two stations. Temperature dependent H-NMR spectra indicated that this process occurred 500 times a second (Stoddart et al., 1992).
Lehn et al. reported a molecular switching device 1→2 shown in FIG. 7 that effects the redox on/off switching of luminescence and combines an electroactive component with a light-emitting center. Both the oxidized and reduced forms are isolatable and stable. The reduced form 2 is luminescent, whereas the oxidized form 1 is quenched. The electrochemical interconversion of the two species was reported to be reversible (Goulle et al., 1993).
Another approach by Lehn et al. featured 1,2-diarylethenes that can undergo reversible ring closure. The open form can be converted almost quantitatively into a closed form by UV light at 365 nm. The reverse process can be effected thermally or photochemically, at 600 nm, as shown in FIG. 8. The process was followed by UV (Giltam et al., 1995).
Sauvage et al. reported electrochemically triggered swinging of a [2]-catenate, taking advantage of the principle of bi-stability, as many systems have before. This is shown in FIG. 9. A transition metal complex has an organic backbone consisting of two interlocking rings. One ring contains two binding site, a bi- and a tridentate, whereas the second ring features only one bidentate coordination site. Thus, by interlocking both rings, a tetradentate and a pentadentate ligation site is formed. Introducing Cu(II) into the system leads to the formation of the pentacoordinate complex, whereas Cu(I) will generate the tetra-coordinate species. The interconversion between both forms of the complex is electrochemically triggered and corresponds to the sliding motion of one ring within the other. The process was followed by UV (Livoreil et al., 1994).
Systems that substantially profit from transition metals in terms of fluorescence behavior have been thoroughly investigated by Fabbrizzi et al. In general, this system consists of a redox-active metal, i.e., Cu, Ni, which is positioned within a tetradentate ligand, which is tethered to a fluorophore. Depending on the metal oxidation state, the ET mechanism is active or quenched (Farrbizzi et al., 1995). This system is shown in FIG. 10.
Recently, systems based on ruthenium, similar to those of Klehn, have been published (Arounaguiri et al., 1999). A porphyrin zinc complex was used for a supramolecular “electro- and proto-photoswitch.” This system is based on almost the same principles as Fabbrizzi's systems. One state is fluorescent, the other is not (Otsuki et al., 1999). Also, a “bicompartmental ligand” found its way into the literature. This system is quite similar to Shanzer's. The ligand possesses a hard and a soft cavity for Fe(III) and Fe(II). During the redox reaction an iron translocation takes place, similar to that described by Shanzer (Belle et al., 1998).
In the tris-([2-pyridyl]methyl)amine (TPA) family of ligands, a distinct twist of the pyridine rings with respect to the central axis of the molecules was observed. That is, the planes of the pyridine rings were always tilted with respect to the central axis of the structure. In crystals, both enantiomeric twisted conformations were present, and in solution, conformational enantiomerism was observed due to facile interconversion by C—C bond rotations. It was found that it was possible to bias the helical twist in TPA by incorporating a substituent in one of the CH2 groups of the ligand, creating a chiral center (cf. FIG. 11, 1). This chiral center dictates the handedness of the propeller-like helicity of the pyridine moieties in compounds 1 and 3. The compounds were characterized in the solid state by X-ray crystallography and in solution by circular dichroism (CD) (Canary et al., 1995). Similar observations were made for all of the compounds 1–4 (Canary et al., 1998). There were remarkably large amplitudes for the CD spectra of the ligands, and it was found that the origin of the signal could be the quinoline rings in solution (Castagnetto et al., 1997). Complexes of several metal ions were studied; it was found that only metals that bind the ligand in a trigonal bipyramidal coordination geometry (ZnII, CuII) gave the propeller twist, and hence the large CD amplitudes. Metal ions forming octahedral complexes with 3 (FeII, CdII) gave a different ligand conformation (Cs), no twist, and low amplitude CD spectra.
The conformation-dependent chiroptical properties were applied to the development of a “smart” chemosensor for metal ions (Castagnetto et al., 1998). As noted above, only certain metal ions give enhanced CD signal amplitude upon binding to 3. It was found that ZnII and CdII give large enhancements in fluorescence intensity of the ligand, but that diamagnetic ions FeII and CuII do not. Thus, it was possible to show that 3 can give two signals, CD and fluorescence, and that these taken together result in the ability to discriminate between four possible metal ions, as shown in FIG. 12. To demonstrate a more practical and biorelevant application, it was shown that it was possible to generate calibration curves for detecting varying amounts of ZnII and CdII, and that a measurement of CD and fluorescence intensities of a single sample could give the concentrations of both species. Another possible application of this technology would be in the determination of ZnII levels in cells by microscopy enhanced with fluorescence-detected CE affording a low background compared with isotropic fluorescence measurements.
The present inventors also developed a method for determination of the absolute configuration of primary amines (Zahn et al., 1999 1). From the previous discussion, one can see that the sequence of the two Cotton effects in the CD spectra reveals the absolute configuration of the chiral atom in the tripod ligand. This correlation also holds true for ZnII complexes of a series of substituted quinoline compounds that were prepared with different substituents as shown in FIG. 13. Only compound 7 fails to give exciton-coupled CD spectra, probably due to the isopropyl group being so sterically encumbering that it distorts the complex. The ligands are prepared from primary amines; thus, derivatization of the primary amines by nucleophilic substitution of bromomethylquinoline (the method used in the synthesis) could be considered a method of absolute configuration determination of the amines. The present inventors have also now shown that the same approach can be used to determine the absolute configuration of α-amino acids and β-amino alcohols. Thus, reaction of an analyte with bromomethylquinoline on as small a scale as 3 mg, followed by addition of Cu(ClO4)2 and NH4SCN, affords a CD spectrum that reveals the absolute configuration of the analyte.
The Cu(I) and Cu(II) complexes of the tripod ligands displayed varying degrees of “twistedness”. (Chuang et al., 1995). This and the fact that the amplitude of the CD spectra of the chiral complexes was strongly dependent on the dihedral angle between the chromophores led the present inventors to investigate the development of a redox-write, CD-read molecular switch. The idea was that the twisting and untwisting of the trisquinoline ligand shown in FIG. 12 would give variable CD intensities. The compound was prepared and its chiroptical properties determined; the observed CD amplitude variation was even larger than predicted due to one quinoline arm decoordinating in the Cu(I) complex in the presence of SCN− counterion. (Zahn et al., 1998). The data for this molecular switch were exceptionally gratifying in terms of the magnitude of the change and the reversibility of the spectra using solution techniques. Additionally, in collaboration with Prof. Gottarelli at the University of Bologna, the present inventors have found that the trisquinoline compounds can be added to nematic liquid crystalline phases to induce cholesteric phases.